Query local Parquet files with DuckDB and return SQL statements.
npx skills add https://github.com/jin-bo/agentao --skill duckdb-analyst
Please help me install this Agent Skill. Skill: duckdb-analyst Source: https://github.com/jin-bo/agentao/tree/main/examples/data-workbench/.agentao/skills/duckdb-analyst Command: npx skills add https://github.com/jin-bo/agentao --skill duckdb-analyst
This Skill eliminates the hassle of manually querying Parquet data files, ensuring you always have full visibility into the exact SQL logic used to generate every analytical result, with no hidden or unverified query steps.
Use the duckdb-analyst skill to calculate the total revenue per product from the sales.parquet file in the ./data directory.
High-intent search queries and answers about installing and using this skill.
You can query Parquet data files with SQL directly using the DuckDB query engine. This approach eliminates the need for a separate database instance, allowing you to run fast analytical queries for data exploration and aggregation immediately.
Yes, full SQL transparency is provided for all generated results. Every query result is paired with the exact SQL statement executed, enabling you to audit, verify, or manually modify the analytical query logic as needed.
Ad-hoc data exploration and profiling on local Parquet files is performed by applying analytical queries to structured datasets. This includes tasks like sales analytics, log analysis, and trend identification directly within the read-only ./data directory.
Yes, analytical queries enforce a default 1000-row limit on outputs. Additionally, destructive SQL operations such as DELETE, UPDATE, and DROP are prohibited to ensure safe, read-only data exploration on your Parquet files.
The best way to calculate total revenue and identify trends from sales Parquet files is using an analytical query engine. You can aggregate total revenue per product, filter top-performing regions, or identify year-over-year trends without writing boilerplate data loading code.
No, you do not need to manually load local Parquet data before running analytical queries. The DuckDB query engine reads Parquet files directly from the read-only ./data directory, bypassing the need for boilerplate data loading code.
starlake-aiConvert Parquet files into CSV format for batch ETL pipelines.
jaimedhenriquesValidate analyses for methodological errors, accuracy issues, and bias.
Sir-chawakornGenerate static, interactive, and dashboard visualizations from structured datasets.
tondevrelRun in-process analytical SQL queries on Python DataFrames and files.
